Supervisor:
Professor Frede Blaabjerg

Co-Supervisor:
Associate Professor Dao Zhou

Assessment Committee:
Associate Professor Yajuan Guan, AAU Energy (Chair)
Jan Richard Svensson, Hitachi Energy Sweden AB
Professor Vivek Agarwal, Indian Institute of Technology

Moderator:
Assistant Professor Subham Sahoo

Abstract:

System inertia and voltage stability are two main characteristics of the electrical power system. In traditional power systems, the system inertia depends on the stored energy in the rotating masses of the traditional synchronous generators (SGs). However, as the penetration of wind power generation increases, numerous SGs will be replaced by converter-interfaced generators (CIGs). Thus, the system inertia will be reduced significantly, which may threaten the frequency stability. Moreover, the long distance between the wind farm and the main grid usually causes a large transmission impedance and a low short-circuit ratio (SCR). In the low-SCR grid, maintaining voltage stability is a challenge. Thus, the power grid with a high percentage of wind energy tends to encounter voltage and frequency instability issues. Therefore, this Ph.D. project focuses on developing improved control solutions to enhance the voltage stability and frequency stability of wind generation systems connected to weak grids with low SCR and low inertia.

Typically, grid-connected inverters in wind generation systems need a phase-locked loop (PLL) for grid synchronization, which helps the inverter to properly inject power into the grid. This PLL-based control method is usually called grid-following (GFL) control method, which is the first research target of the thesis. Specifically, the small-signal state-space model and impedance model of GFL inverters are built to analyze the voltage instability mechanism. According to the small-signal impedance model, it is found that the positive feedback introduced by the PLL can lead to negative resistance characteristics, which are identified as the root cause of the instability. In order to mitigate the positive feedback and weaken the negative resistance characteristic, a double-PLL-based impedance reshaping control method is proposed, which can not only extend the small-signal stability range, but also improve the dynamic performance of the PLL. Thus, the voltage instability issue of the conventional GFL inverters in weak grids can be solved by using the proposed control method.

Secondly, grid-forming (GFM) control methods become the next research focus, because GFM inverters are necessary for low-inertia grids to improve the frequency stability. Considering that GFM control schemes have not been unified or standardized yet, three commonly used GFM control schemes are compared initially. It is found that a virtual-admittance-based GFM control scheme has a wider stability range than a virtual-impedance-based scheme and the typical dual-inner-loop control scheme. So, the virtual-admittance-based GFM control scheme is chosen to be used in this thesis. Then, the parameter design method for the virtual admittance is studied and a simple design method is proposed. By using the proposed parameter design method, GFM inverters are able to be stable in normal grid conditions with a wide range of SCR. Moreover, it is found that GFM inverters with a conventional current reference limiting method tend to be unstable under large grid disturbances. To solve this problem, a power-angle-based adaptive overcurrent protection method is proposed, which can maintain the stability of GFM inverters under large grid voltage or frequency disturbances. Thus, with the proposed overcurrent protection method, GFM inverters can operate stably even under large grid disturbances.

Afterwards, to study stability issues of wind generators, small-signal models of Type-3 and Type-4 wind generation systems are built. For Type-4 wind generation systems, only considering the grid-side converter (GSC) is enough to derive the small-signal impedance model and perform the stability analysis. However, for Type-3 wind generation systems, it is difficult to obtain the full-order impedance model of the doubly-fed induction generator (DFIG) system due to both dc-side and ac-side couplings between the GSC and the rotor-side converter (RSC). To address this difficulty, a two-port-network-based decoupled impedance modeling method is proposed, which is able to decouple the ac-side coupling so that the RSC, GSC, and dc-link coupling can be modeled separately. By using the proposed modeling method, the impact of the dc-link coupling in the DFIG system with either GFL or GFM control can be analyzed quantitatively. Thus, according to small-signal models of PMSG and DFIG systems, the small-signal stability of grid-connected wind generators can be analyzed accurately.

Then, the proposed improved GFL and GFM control methods are applied to two paralleled permanent magnet synchronous generator (PMSG)-based Type-4 wind generators, and it is found that the stability and transient performance of GFL and GFM Type-4 wind generators can be improved. So, the requirements defined in the Grid Codes (e.g., the voltage stability, frequency stability, and low-voltage ride-through) are basically satisfied. Finally, a more realistic case study of Anholt Offshore Wind Power Plant is carried out to show the effectiveness of the proposed GFL and GFM control solutions.

The outcomes of this Ph.D. project will be helpful to address stability challenges of power systems with a high percentage of wind power generation.